Vascular malformations of the brain are a rare but important cause of stroke. Clinically, an important subtype is brain arteriovenous malformation (AVM). As-yet unknown defects in angiogenesis and vascular assembly are thought to undergird development of the clinical phenotype. Understanding the signaling which results in proper CNS vascular development and integrity is the likeliest approach to developing relevant therapies to prevent hemorrhage in these conditions. A unifying theme of this proposal is a vertically integrated program that can relate clinical observations to various aspects of the underlying genetic and cell-to-cell signaling abnormalities. Project 1 (Young) is a clinical investigation that will identify novel candidate genetic variants associated with AVM susceptibility using genome-wide association methods, therefore encompassing project-related signaling pathways. Project 2 (Hashimoto) addresses macrovascular remodeling due to matrix metalloprotease (MMP) activity and inflammatory activity. Project 3 (Boudreau) concerns homeobox genes, which are master regulatory mechanisms in the regulation of extracellular matrix and angiogenesis;this project focuses on the anti-angiogenic properties of HOX A5, which we have found to be deficient in human BAVM tissue. Project 4 (Nishimura) investigates the role of astrocyte-endothelial cell interactions in a key signaling pathway for cerebrovascular homeostasis- integrin-mediated control of TGF-?;TGF-? signaling is implicated in the only known heritable form of AVM, i.e., hereditary hemorrhagic telangectasias. The three cores serve all projects. The Administrative Core A (Young) coordinates PPG activities. The Data Management Core B (McCulloch) serves as the central mechanism for clinical data collection, organized data input and analyses. The Laboratory Core C (Su) furnishes a central laboratory resource for models used in the laboratory project including murine vascular dysplasia, flow loading of large arteries, and human-to-rodent tissue transplant;and serves as a central human surgical specimen tissue bank. At the present time, AVM treatment is extirpative and entails relatively high costs with significant risks. Improved mechanistic insight into the pathophysiology of the disease will facilitate development of novel therapies and biomarkers for a disease that currently has no specific medical therapy. This program represents a unique coupling of clinical and basic investigators, along with a unique, large clinical database and tissue bank, to address a complex and important clinical problem.